The EPRI Magnet Motor

Cross-section of the EPRI magnet motor, from US Patent 5,304,882.
Note the flat permanent magnets under the rotor pole-pairs, and the stator of typical
switched reluctance design, similar to the Takahashi Self-Generating Motor.

Permanent magnets in a switched reluctance motor

An early "significant patent" [Ref 1] for a switched reluctance motor with permanent magnets is US Patent 5,304,882 of 19 April 1994, issued to Thomas Lipo and Yuefeng Liao of the prestigious Electric Power Research Institute (EPRI). According to the Wikipedia entry on EPRI, "The institute's research and development program spans every aspect of generation, environmental protection, power delivery, retail use, and power markets. EPRI provides services to more than 1000 energy-related organizations in 40 countries. It has more than 900 patents to its credit."

Was it ever built?

While this patent 5,304,882 claims many very important improvements over previous designs, (including "higher efficiency") I can find no evidence that the patented machine (shown above) was ever actually built — it seems to have been entirely a series of computer simulations. 

Free energy is implied

The most provocative claim was that the invention could produce between √2 and 2√2 times as much torque as a comparably-sized conventional machine having the same winding heat dissipation and degree of magnetic saturation. Since energy output is directly proportional to torque, through a given angle of rotation, this implies an energy output of more than 1.4 to 2.8 times that of the conventional machine for the same current (and hence presumably for the same energy input). Since a conventional switched reluctance motor can achieve 94% efficiency, this claim, supported by figures from the patent shown below, implies that this EPRI motor could quite easily deliver free energy, although nothing explicit was said about that.

Figures from US Patent 5,304,882, showing energy conversion loops of magnetic flux linkage (here labelled λ, but more usually ψ) vs stator coil current.

Note especially the parallelogram-shaped energy loops of Figures 8A and 8B, showing operation in the second
quadrant as well as the first, giving a much larger total energy output than could be achieved with an equivalent
conventional switched reluctance motor (represented by the smaller cross-hatched loops). 

Notable points

The following points about the EPRI motor are worth noting:

-  Although the patent is titled "Variable reluctance motors with permanent magnet excitation," the authors state that "...The essential feature is that the permanent magnets see a constant reluctance flux path at all times, that is, the total overlapping stator/rotor pole areas must remain constant." (My emphasis).

-  Figures 8A and 8B above strongly suggest that the motor is able to use current reversal to achieve magnetic repulsion as well as attraction.

-  Considering all of the major improvements claimed for their patented motor, it is surprising that EPRI apparently did not build any version of it as a physical prototype — at least no such prototype has been publicised, to my knowledge. It is also surprising that more than twenty years later, manufacturers still seem uninterested in incorporating permanent magnets into their commercially available switched reluctance motors in this way, in order to achieve anything like the energy conversion loops shown above.

Are these energy conversion loops too obvious?

In two subsequent patents, also on variable reluctance machines incorporating permanent magnets, US Patent 5,672,925, and WO9418855, the same EPRI inventors did not provide any energy conversion loops, whereas by analogy with US Patent 5,304,882 they would have been expected.

On espacenet, I found one more associated patent from these inventors, US Patent 5,825,112. This patent does have the provocative large "two quadrant" energy conversion loops similar to those shown above, and again claims to produce torque of between √2 and 2√2 times that of a conventional motor. 

Curiously, I found on espacenet that I could easily see the mosaics (patent images) from the two patents that did not have any energy conversion loops, whereas I could not see the mosaics from those (US Patents 5,304,882 and 5,825,112) that did have them.

I've mentioned before that although espacenet remains my favourite site for patent searches, I do occasionally have a few problems with it, so I don't really think this is an attempt at suppression! (If it is, it's a very feeble one — the patent images are easily downloadable and visible in all cases from the "original document" tab on their patent webpages).

Questions

As usual, there are plenty of unanswered questions concerning this EPRI work, in particular the motor of US Patent 5,304,882. I'll confine myself to just a few important ones:—

(for EPRI):—

-  Was a physical prototype ever built? If not, why not, in view of all the claimed advantages? If it was, what were its test results?

-  Can any interpretation be put on the large parallelogram-shaped energy loops of Figures 8A and 8B, other than that they show over-unity energy output? If so, what?

-  Why were energy loops omitted for US Patent 5,672,925, and WO9418855?

(for switched-reluctance motor manufacturers):—

- Why are you apparently not interested in incorporating permanent magnets into your motors and controlling them in the way shown in this EPRI work, so that their energy loops can extend into the second as well as the first quadrant, to achieve the far better performance claimed by EPRI (and also by Yasunori Takahashi, with his "self-generating" motor of very similar design, which he explicitly claimed to be a free energy motor)?

Reference

1. T J E Miller, Electronic Control of Switched Reluctance Machines, Newnes, p32.